Ecology, 101(8), 2020, e03086 © 2020 by the Ecological Society of America Reproductive phenology across the lunar cycle: parental decisions, offspring responses, and consequences for reef fish 1,6 2 3 4 JEFFREY S. SHIMA , CRAIG W. O SENBERG , SUZANNE H. ALONZO, ERIK G. NOONBURG, 1 5 PAULINE MITTERWALLNER, AND STEPHEN E. SWEARER 1School of Biological Sciences, Victoria University of Wellington, Wellington 6140 New Zealand 2Odum School of Ecology, University of Georgia, 140 East Green Street, Athens, Georgia 30602 USA 3Department of Ecology and Evolutionary Biology, University of California at Santa Cruz, Santa Cruz, California, USA 4Biological Sciences, Florida Atlantic University, Davie, Florida 33314 USA 5School of Biosciences, University of Melbourne, Melbourne, Victoria 3010 Australia Citation: Shima, J. S., C. W. Osenberg, S. H. Alonzo, E. G. Noonburg, P. Mitterwallner, and S. E. Swearer. 2020. Reproductive phenology across the lunar cycle: parental decisions, offspring responses, and conse- quences for reef fish. Ecology 101(8):e03086. 10.1002/ecy.3086 Abstract. Most organisms reproduce in a dynamic environment, and life-history theory predicts that this can favor the evolution of strategies that capitalize on good times and avoid bad times. When offspring experience these environmental changes, fitness can depend strongly upon environmental conditions at birth and at later life stages. Consequently, fitness will be influenced by the reproductive decisions of parents (i.e., birth date effects) and developmental decisions (e.g., adaptive plasticity) of their offspring. We explored the consequences of these decisions using a highly iteroparous coral reef fish (the sixbar wrasse, Thalassoma hardwicke) and in a system where both parental and offspring environments vary with the lunar cycle. We tested the hypotheses that (1) reproductive patterns and offspring survival vary across the lunar cycle and (2) offspring exhibit adaptive plasticity in development time. We evaluated temporal variation in egg production from February to June 2017, and corresponding larval develop- mental histories (inferred from otolith microstructure) of successful settlers and surviving juve- niles that were spawned during that same period. We documented lunar-cyclic variation in egg production (most eggs were spawned at the new moon). This pattern was at odds with the dis- tribution of birth dates of settlers and surviving juveniles—most individuals that successfully survived to settlement and older stages were born during the full moon. Consequently, the probability of survival across the larval stage was greatest for offspring born close to the full moon, when egg production was at its lowest. Offspring also exhibited plasticity in develop- mental duration, adjusting their age at settlement to settle during darker portions of the lunar cycle than expected given their birth date. Offspring born near the new moon tended to be older and larger at settlement, and these traits conveyed a strong fitness advantage (i.e., a car- ryover effect) through to adulthood. We speculate that these effects (1) are shaped by a dynamic landscape of risk and reward determined by moonlight, which differentially influ- ences adults and offspring, and (2) can explain the evolution of extreme iteroparity in sixbars. Key words: adaptive plasticity; carryover effect; coral reef fish; ecoevolutionary feedback; life history; lunar periodicity; otolith microstructure; recruitment; reproductive output; seasonality; selection; settlement. traits (e.g., when and where to reproduce; Johannes INTRODUCTION 1978, Thomas et al. 2001). Even when components of Most organisms live in a dynamic environment, where environmental variation are predictable (i.e., seasonal- environmental conditions (e.g., food, shelter, predators) ity), selection can favor some form of trait variation or vary through time (Fretwell 1972). This implies a series associated strategy that enables flexibility (e.g., adaptive of “good” and “bad” times for activities such as forag- plasticity; Ghalambor et al. 2007, Shima et al. 2018). ing, migration, and reproduction. Temporal variation in Reproductive phenology is a particularly good exam- the environment contributes to variation in demographic ple of this. Many organisms (e.g., plants [Wheeler et al. rates, which affects population dynamics (Reid et al. 2015], insects [Forrest and Thompson 2011], birds 2018) and can influence the evolution of life-history [Dunn and Winkler 1999], mammals [Middleton et al. 2013], amphibians [Gibbs and Breisch 2001], and marine fishes [Genner et al. 2010]) reproduce seasonally, and at Manuscript received 13 December 2019; revised 19 March 2020; accepted 1 April 2020. Corresponding Editor: Lorenzo times that are likely to increase lifetime expected repro- Ciannelli. ductive success. For semelparous organisms (e.g., many 6 E-mail: [email protected] annual plants, Pacific salmon), reproductive timing Article e03086; page 1 Article e03086; page 2 JEFFREY S. SHIMA ET AL. Ecology, Vol. 101, No. 8 should be driven primarily by strategies that increase variation in nocturnal illumination, which has the poten- offspring success. For organisms that reproduce more tial to strongly shape the growth and survival of early than once, reproductive timing may reflect a compro- life-history stages of reef fishes in the pelagic environ- mise between strategies that increase the survival and ment (Hernandez-Leon 2008, Shima and Swearer 2019). expected reproductive success of individual offspring Here we focus on a highly iteroparous coral reef fish, and those that increase adult survival and future repro- the sixbar wrasse (Thalassoma hardwicke). We compare duction (Warner 1998, Wadgymar et al. 2017). lunar variation in reproductive output to birth dates of Environmental predictability and the timescale of offspring that survive to settle back onto the reef (which environmental changes can greatly influence patterns of we reconstruct from otolith microstructure). We estimate reproduction. For example, many marine fishes exhibit relative survival of offspring born across the lunar cycle seasonal reproduction in response to seasonal plankton to test the hypotheses that (1) survival varies as a func- blooms (i.e., larval food availability), with the degree of tion of lunar birth date and (2) offspring exhibit adap- temporal concordance between spawning and food pro- tive plasticity in development time (i.e., as predicted by duction affecting offspring survival (Cushing 1975, Shima et al. 2018). More specifically, we predict that Beaugrand et al. 2003). Studies conducted at finer tem- selection should favor birth dates and/or adjustments to poral scales demonstrate how “sweepstakes” events (e.g., larval developmental durations that facilitate settlement disproportionate contributions from a small number of back to the reef on the new moon. Our results illustrate individuals or spawning bouts) may also be important in how an interaction among early life-history traits— these systems (Hedgecock and Pudovkin 2011, Shima specifically, birth dates imposed by parents and plastic- and Swearer 2016). The prevalence of sweepstakes ity in developmental duration by offspring—may drive recruitment suggests that environmental unpredictability phenotypic variation that shapes selection through to may impede optimization of spawning times, and this adulthood. may partly account for a preponderance of extreme iteroparity among many fishes (Johannes 1978, Claydon METHODS et al. 2014). Coral reefs are less seasonal in terms of water temper- Study system ature and pelagic larval food availability relative to habi- tats in higher latitudes. Nonetheless, important cyclic We estimated temporal variation in reproductive out- patterns may be imposed by the lunar cycle, which can put, birth dates, and larval traits of recently settled six- synchronize reproductive activities (Foster et al. 2018), bar wrasse (Thalassoma hardwicke) sampled from the transport food, eggs, and/or larvae (via tides: Connell northern lagoon of Moorea, French Polynesia. Sixbars 1961, Sponaugle and Cowen 1996, Forward and Tanker- are coral reef fish, commonly found on shallow fringing sley 2001), and affect the outcome of predator–prey reefs and lagoons throughout much of the Indo-Pacific interactions (via nocturnal illumination: Kronfeld-Schor region. They are protogynous hermaphrodites, and et al. 2013, Prugh and Golden 2014, Palmer et al. 2017). spawn pelagic eggs at sites that are typically located at Many coral reef fishes spawn at particular times of the reef edges, near passes (i.e., gaps in a barrier reef crest). lunar cycle (Johannes 1978, Claydon et al. 2014), and a Spawning activity is generally greatest between 14:00 given pattern of reproductive periodicity may maximize and 16:00 on Moorea, at times when offshore water flow adult survival and/or offspring performance (Robertson is strongest (P. Mitterwallner, unpublished data). These et al. 1990). The larval stages of many of these species spawning patterns may serve to maximize offshore trans- also settle back to the reef during particular moon port of eggs (i.e., away from high concentrations of reef- stages; most commonly, settlement occurs at night, and based egg predators; Johannes 1978, Robertson et al. close to the new moon (Dufour and Galzin 1993, Ran- 1990). kin and Sponaugle 2014, Shima et al. 2018). Settlement Spawning occurs most days of the lunar cycle and near the new moon is generally assumed to be an adap- throughout